Assessment of Bone Marrow Oxygenation Based on T2* and T2 Changes Following Oxygen Inhalation

Background Osteoporosis affects 1 in 4 women and 1 in 8 men over 50 years and its burden on healthcare resources is enormous. Despite the risk factors and biological processes leading to bone loss have been well studied, little is known about the mechanisms that initiate bone loss in the first place. Osteocytes, marrow adipocytes and haematopoetic cell lines share a common precursor in the marrow mesenchymal stem cell (MSC) which can variably differentiate along osteoblastic, adipocytic or haematopoietic cell lines. The level of oxygenation within bone marrow may affect the pattern of MSC differentiation. On the other hand, a shift in the pattern of MSC differentiation might be observed indirectly by a change in the demand for oxygen in the bone marrow. Therefore, development of an MR-based technique to study marrow oxygenation non-invasively might be helpful to further our understanding on marrow physiology and bone formation. Introduction MR relaxometry (T2* and T2) techniques have been shown to be useful in osteoporosis studies (1). Patients with osteoporosis have prolonged marrow T2* due to reduced magnetic field inhomogeneities in less dense trabecular bone (2). Accumulation of fatty marrow in osteoporotic bone is believed to increase the marrow T2 of subjects with reduced bone mineral density (BMD) (2,3). Deoxyhemoglobin is paramagnetic and breathing carbogen or pure oxygen lowers deoxyhemoglobin concentration and increases both T2* and T2 of water in blood and in the tissue surrounding blood vessels (4,5). Our aim in this study was to verify whether oxygen inhalation has a measurable effect on bone marrow T2* and T2 relaxation times. Material and Methods Seven healthy volunteers (4 males and 3 females; mean age 35 years) were examined in the supine position on a 3.0T scanner (Achieva, Philips Healthcare, Best, The Netherlands) using a standard spine coil. Air or 100% oxygen was delivered via a tight-seal full-face mask (Mirage NV, ResMed, Sydney, Australia) at a rate of 15 L/minute. A 10 mm thick axial image was acquired from each vertebral body L3, L4 and L5 for relaxation measurements before (T2* air and T2 air) and after 5 minutes of oxygen inhalation (T2* oxygen and T2 oxygen). A multi-echo fast field echo sequence (TR/TE/δTE 93/1.9/1.5 ms; 12 echoes; in-plane pixel size 1.5×1.2 mm) was used to measure T2*. For T2 measurement, a multi-echo spin-echo sequence was employed (TR/TE 990/n×20 ms; n = 1-12; in-plane pixel size 1.5×1.2 mm). A pixelby-pixel T2* map (Fig. 1) and T2 map (Fig. 2) were obtained. For a given vertebral body, the same region of interest was drawn on the maps to obtain mean T2* air, T2* oxygen, T2 air and T2 oxygen. Paired t test was used to evaluate differences T2* air vs. T2* oxygen and T2 air vs. T2 oxygen with differences considered significant at P < 0.05. Results Twenty-one pairs of T2* and T2 values before and after breathing oxygen were obtained from the 3 vertebral bodies of all 7 subjects. Table below summarizes the results after breathing air and pure oxygen for 5 minutes. On average, T2* increased by 11.7% compared to 7.9% for T2. Pair t-test showed that there were significant differences in T2* (p = 0.002) and T2 (p < 0.0001) values when breathing air or oxygen.